Biol Res 25: 41-49(1992) The evolutionary origin of the ...

Biol Res 25: 41-49(1992)

41

The evolutionary origin of the mammalian cerebral cortex

FRANCISCO ABOITIZ*

Neuroscience Program, Brain Research Institute, University of California, Los Angeles, CA 90024-1761, USA

The origin of the mammalian neocortex in usually considered as an improvement in the structure of the brain. Alternatively, I suggest that the mammalian neocortex arose as a consequence of contingent adaptations in which there was no specific selection for more elaborate cognitive abilities. In primitive mammals, the adaptation to nocturnal life produced a reduction of the optic tectum (superior colliculus). In addition, the development of the olfactory system triggered the development of the cerebral cortex. It is proposed that , since both the optic tectum and the cerebral cortex are laminar structures, the growing cortex replaced the tectum in many integratory functions. When mammals reinvaded diurnal niches, the optic tectum did not redevelop, and the cerebral cortex remained the main integratory and perceptual system. This is a case of irreversible reduction of an organ. In reptiles and especially in birds, although there was also an increase in brain size (associated with higher cognitive capacities), the optic tectum grew in size and complexity and the forebrain grew largely as a nonlaminar structure (except the Wulst in birds). Therefore, the origin of the cerebral cortex resulted from the combination of adaptations to nocturnality and the development of olfactory-driven behavior, and its origin is not directly related to higher cognitive capacities.

ABBREVIATIONS

AR BG CX DMN DVR EN FC HP

NC OT

PC PT PY T

w

Archicortex Basal ganglia

Cerebral cortex Dorsomedial nucleus of thalamus Dorsal ventricular ridge Entorhinal cortex Frontal cortex Hippocampus

Neocortex Optic tectum / superior colliculus

Paleocortex Pretectum Pyriform cortex Thalamus Wulst

INTRODUCTION

Although all vertebrates share a common plan of brain organization, there is also diversity in more detailed aspects of brain structure (Sarnat and Netsky, 1981; Northcutt, 1981; Ulinski, 1990b). Mammals are characterized by the development of a large, laminar

neocortex that receives sensory projections from the thalamus, and a very reduced optic tectum (the superior colliculus, a laminar structure) that receives direct retinal projec tions (Fig. 1A). On the other hand, reptiles have developed a nuclear structure, the dor sal ventricular ridge (DVR), which is similar to the basal ganglia in terms of superficial histology and location (Fig. IB). DVR re ceives most sensory thalamic projections and is considered homologous to parts of mam malian neocortex at least in terms of connec tivity (Northcutt, 1981; Ulinski, 1983, 1990b). Reptiles have also retained a well developed laminar optic tectum, receiving an important proportion of the visual projections from the retina. Reptiles and amphibians have a primi tive cerebral cortex, but it is not nearly as developed as it is in mammals (Ulinski, 1990a). Birds (Fig. 1C) have a slightly modi fied reptilian plan of brain organization (Ulinski, 1990b; Ulinski and Margoliash, 1990), with an especially large optic tectum and a reduced cerebral cortex (lateral, medial and dorsomedial). Birds have also developed a laminar structure related to the cerebral cortex (the Wulst), and their DVR is thick-

New address: Departamento de Fisiolog?a y Biof?sica, Facultad de Medicina, Sede Norte, Universidad de Chile, Casilla 70005, Santiago 7, Chile.

42

Biol Res 25: 41-49 (1992)

A

ened in relation to reptiles and retains its nu

clear architecture.

For several authors, the origin of the mam

malian neocortex has been implicitly or ex

plicitly considered as a structural improve

ment over other types of brain organization

(Papez, 1929; Jackson, 1931; MacLean, 1973;

Brown, 1967, 1991; Glezer et al, 1988;

Allman, 1990), despite the fact that at com

parable levels of encephalization, birds and

mammals do not show significant differences

in cognitive abilities (Hodos, 1970; Walker,

1983). Recently, Airman (1990) has proposed

that the increased development of laminar

structures in the brains of birds (optic tectum

B

and Wulst) and of mammals (neocortex) are

associated with exploratory behavior that de

velops with homeothermy. This hypothesis

may explain some facts such as the larger

brain size in these two groups as compared to

reptiles. However, it does not account for the

differences in brain structure between birds

and mammals, i.e., the striking reduction of

the optic tectum (being a laminar structure)

in mammals, and the fact that in birds DVR

remains largely a nuclear structure.

In this paper, I suggest that the emergence of the mammalian cortex can be attributed more to historical circumstances that in this case resulted in an iiTeversible change in or ganization, rather than to a selective advantage in perceptual and cognitive abilities. The hy pothesis proposes an explanation of the re duction of the optic tectum (superior colliculus) in mammals, and of the mainte nance of a large, nuclear DVR in birds.

Fig. 1: Highly simplified diagrams of projections from the retina and the basal ganglia in mammals (A), reptiles (B) and birds (C). Basal ganglia projections to the optic tectum via the substantia nigra are not shown. BG, basal ganglia; CX, cerebral cortex; DVR, dorsal ventricular ridge; T, thalamus; OT, optic tectum (superior colliculus in mammals); PT, pretectum; W, Wulst.

The origin of the mammalian neocortex

Although it may have been in an incipient state in mammal-like reptiles (Quiroga, 1980), the mammalian neocortex developed among the first true mammals (Jerison, 1973, 1990; Hopson, 1979; Quiroga, 1980; Kemp, 1982), which were small sized, nocturnal animals, superficially similar to present-day rodents and insectivores. It has been proposed that, as an adaptation to nocturnal life, mesozoic mammals did not have a strong sense of vi sion, resulting in a reduction of the optic tectum (superior colliculus) (Jerison, 1973).

Concomitantly, mesozoic mammals devel oped an elaborated olfactory system (Jerison,

Biol Res 25: 41-49 (1992)

43

1973; Kemp, 1982; Lynch, 1990). The olfac tory bulbs project directly to the olfactory cortex in both reptiles and mammals (Haberly, 1990). It is likely that in protomammals, paleo and archicortical structures (receiving primary and secondary olfactory projections from the olfactory bulb; Fig. 2) became more developed in response to increasing olfactory develop ment (Lynch, 1990; Ulinski, 1990a). It has been suggested that, in evolution, the elabo ration of sensory systems may produce an expansion of the central regions receiving these projections (Woolsey and Van der Loos, 1970; Welker and Van der Loos, 1986).

The secondary olfactory afferents include, among other structures, the dorsomedial nu cleus (DMN) of the thalamus, that projects to the frontal cortex, and true frontal cortex itself (Haberly and Price, 1978a, b; Haberly, 1990; Fig. 2). It is interesting that these projections seem to be better developed in primitive mammals. The DMN of the opossum (marsu pial) has much denser olfactory projections than is the case in placental mammals (Benjamin et ai, 1982). In addition, in the echidna (monotreme), which has a highly de veloped olfactory system, the projection of the DMN occupies a vast area of the anterior neocortex (Welker and Lende, 1980). This evidence has in part led Lynch (1990) to pro pose that the increasing olfactory projections had a dominant role in the evolutionary emergence and development of the neocortex.

Mesozoic mammals also developed a sen sitive audition, partly based on the acquisi-

HP

H

I r l?f 1

PJV

I PA

i

zzn 0

Fig. 2: Scheme of some olfactory projections in the mam malian brain. AR, archicortex; DMN, dorsomedial nucleus of the thalamus; EN, entorhinal cortex; FC, frontal cortex; HP, hippocampus; NC, neocortex; OB, olfactory bulb; PA, paleocortex; PY, pyriform cortex. (Modified from Lynch, 1990).

tion of a dentate-squamosal jaw articulation, and the consequent full incorporation of the ear ossicles into the auditory system (Jerison, 1973; Kemp, 1985). In addition, there may have been increased somatosensory sensitiv ity associated with the loss of scales in the protomammalian lineage. Apparently, the de velopment of these sensory modalities was associated to the growth of their telencephalic projections.

I suggest that in mammals, an additional factor favoring the growth of thalamo-cortical sensory projections (somatosensorial, au ditory and visual) was the establishment of associative connections between these modalities and the olfactory system, which served in the cognitive elaboration of sensory stimuli. The olfactory projection forms ex tensive associations with the neocortex in the hippocampus and other cortical areas (Lynch, 1990). These associations may have contrib uted to fuse different perceptual modalities into multisensorial "objects".

These were not the only changes that oc curred. In reptiles, the basal ganglia (involved in motor functions) send two major indirect projections to the optic tectum, one via the pretectum (see Fig. IB) and the other via the substantia nigra (Reiner et ai, 1984; Ulinski, 1986; Medina and Smeets, 1991). Besides its optic function, the optic tectum is a principal center for the integration of complex sensory information and motor signals (Ulinski, 1983, 1986). In mammals, similarly to retinal pro jections, the output of the basal ganglia to the pretectum and then to the optic tectum is de viated to the thalamus (motor and intralaminar nuclei; Reiner et ai, 1984; Ulinski, 1986; Brauth, 1990; Fig. 1A). In reptiles there may be some projections from the basal ganglia to the thalamus (Parent, 1986), but in any case they do not seem as prominent as in mammals (Ulinski, 1990b; Brauth, 1990). Additionally, in mammals direct, descending connections from the cerebral cortex reach the reticular formation and the spinal cord, while in reptiles they never go below the basal ganglia (Ulinski, 1986). Also, in mammals there is an increase of the ascending cerebellothalamic projections (Ulinski, 1986). Thus, the cerebral cortex of mammals engages in close relations with the basal ganglia, cerebellum, motor brainstem and spinal cord. This prob-

44

Biol Res 25: 41-49 (1992)

ably produces a more sophisticated motor control for performing complex movements with the jaws and limbs, and may also be associated with an increased somatosensory sensitivity.

In summary, in mammals the sensory and motor projections to the tectum which are characteristic of reptiles and amphibians were in large part shifted to the cerebral cortex via the thalamus, with the simultaneous reduction of the tectum. This may have largely been a consequence of reduction of the visual system and increasing olfactory development, with an expansion of the olfactory cortex and the consequent growth of neocortical areas re ceiving olfactory projections. Furthermore, the neocortex may have grown by virtue of in creasing associative connections between ol factory projections and thalamocortical sen sory projections (somatosensory, auditory and visual). The cerebral cortex also began to be involved in the control of complex motor ac tions, with the concomitant development of close associations with the motor systems of the brain.

All this process was accompanied by a dramatic increase in the size of the brain, produced as a consequence of the expansion of the neocortex (thalamo-telencephalic pro jections), and related to the elaboration of cognitive abilities (Jerison, 1973). However, the origin of the neocortex was not a direct consequence of selection for higher cognitive capacities, but rather resulted from adaptations to specific circumstances (reduction of the visual system and olfactory driven behavior). In other words, although the larger brain size of mammals may be related to an increase in cognitive capacities over reptiles, the fact that the particular structure that increased its size and complexity was the cerebral cortex was entirely circumstantial. (Birds increased their brain size and developed their cognitive abilities to a level comparable to mammals by using a different strategy.)

An irreversible shift of functions

I propose that at this point the cerebral cortex of mammals replaced the reduced tectum in many of its perceptual, integratory and motor functions. Perhaps the development of close links between the neocortex and the motor

systems in the brainstem was a key step in this replacement. This proposal is consistent with the concept of redundancy (degeneracy) in neural systems, where several neuronal populations located in different anatomical structures, may be capable of performing a certain function (Edelman, 1987). In this view, most neural functions are not strictly local ized in a given brain region, but are suscepti ble to be performed in other anatomical loci as well. Although it may be difficult at this point to ascertain the specific functions that were replaced (Ingle, 1973; Stein and Gathier, 1981), I must note that both the optic tectum and the cerebral cortex, being laminar struc tures, are especially well suited for the estab lishment of high-resolution two-dimensional maps of neural projections (Edelman, 1987; Allman, 1990).

After the decline of reptiles, mammals un derwent a major adaptive radiation. They reinvaded the diurnal niches, redeveloped their visual systems and the olfactory system be came diminished in relative importance (Stephan et al, 1970; Stephan, 1983). How ever, the optic tectum did not redevelop ac cordingly. It remained small and the thalamocortical system continued receiving the main sensory projections and controlling complex behavior. In other words, the shift of functions from the tectum to the thalamocortical system that had occurred pre viously turned out to be irreversible in this case. I have stated elsewhere (Aboitiz, 1989) that regressing organs that lose their function or whose function is taken by other organs are likely to be irreversibly lost. An extreme case of this situation is the well-known case of the irreversible loss of gills in tetrapods. Even when some tetrapods returned to live in water, they never reacquired gills because these were not functional in the moment of the reinvasion. In other words, when the his torical path is reversed, the function origi nally played by the atrophied organ will either be played by some other organ or will not exist. In this case, selection will tend to act on a functional organ that performs a similar function, while the original organ will remain reduced.

The evolutionary history of the mammalian superior collicullus (optic tectum) seems to fit this pattern. After the reduction of the optic

Biol Res 25: 41-49 (1992)

45

tectum and the development of the olfactory system, many perceptual and motor functions residing mainly in the optic tectum begun to be performed in the cerebral cortex. Two key steps in this process were probably (i) the development of sensory associations between olfaction and the other senses in the hippocampus and some neocortical areas, and (ii) the elaboration of motor control by the neocortex, that permitted the performance of complex, learned motor actions. When mam mals became diurnal again, the cerebral cor tex was so involved in perception and motor command that the simpler strategy was to modify this working system instead of re building the old visual system.

The further evolution of neocortex

I have proposed (Aboitiz, 1988) that the growth of the mammalian cerebral cortex favored the development of multiple corticocortical mapping systems that provide spe cial properties to the brains of some mam mals, especially primates (see Edelmann, 1987). The cerebral cortex may have been especially well suited for the elaboration of these multiple projections by virtue of its laminar structure (Allman, 1990).

The evolutionary growth of the neocortex has been associated with an increase in neuroblast proliferation, resulting in an in creased number of cortical cells (Rakic, 1988). This increase in cell number is mainly related to the addition of cortical columns (Rakic, 1988). However, across species the number of neurons increases at a slower rate than brain size (Jerison, 1973), thus increasing the rela tive size of the neuropil as brains grow larger. This results in an increased space for synaptic terminals. Furthermore, in phylogeny the cere bral cortex grows at a much faster rate than any other brain structure, accounting for most of the increase in brain size (Hofman, 1990). This produces an increase in cortical volume (mainly an increase in cortical surface) that is not matched by subcortical structures. Consequently, subcortical afferents to the ex panded cerebral cortex find an excess of space to make synapses on. On the other hand, the cortical efferent system (mainly the projec tion to the basal ganglia and reticular forma

tion) may overcrowd its subcortical targets by virtue of its growth.

In the perinatal period, there is an exten sive process of retraction of axon collaterals in the cerebral cortex, based in large part on the competition for synaptic targets (Cowan et al., 1984; Purves and Lichtman, 1985; Innocenti, 1986). This dramatically restricts the exuberant cortical connections of the newborn into the adult pattern. I suggest that in an expanded cerebral cortex, descending cortical projections (to basal ganglia and reticular formation) tend to suffer an exten sive retraction of terminals, due to the overcrowding of their subcortical targets. On the other hand, many of these cortical efferents have been found to send collaterals back to the cerebral cortex both in the new born and the adult (Stanfield et al., 1982; Fisher et al., 1986). In an expanded cerebral cortex, these recurring collaterals may find an excess of space to make synapses on and have a better probability of becoming stabi lized. The net result of this hypothetical pro cess would be an increase in the proportions of cortico-cortical projections that become stabilized, at the expense of cortico-subcortical efferents (Aboitiz, 1988). This may be a starting point for the development of com plex cortico-cortical mapping systems that may result in increased cognitive capacities (Edelman, 1987).

Considering the evolution of the human brain, Geschwind (1964) argued that in man, the cerebral cortex would become more inde pendent of subcortical structures by virtue of increasing the relative extent of cortico-corti cal connections. Besides proposing a devel opmental mechanism for this process, I sug gest that this phenomenon occurs not only in the evolution of the human brain, but as a general epigenetic consequence of the increase in the cerebral cortex in relation to subcortical structures. This phenomenon is most dramatic in man, but occurs as well in other cases of increase of cortical size and complexity.

I must emphasize that these multiple cortico-cortical projection systems may have become significant only late in the history of mammals, especially in animals such as pri mates, elephants and dolphins, and may have not been related to the origin and early evo lution of the mammalian cortex.

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download